99% of the books written, paintings painted, sculptures sculpted, poems penned, plays wrighted, and music composed, are of little or no value in the estimation of the public. Millions of people are making stuff that nobody else either wants or buys. They always have - the internet just makes it visible.
The concept of “value” is pretty simple. Do significant numbers of people like something enough to pay money for it?
If people prefer to pay money for AI generated material, that’s because, at least in terms of value, the AI stuff is better. If human creatives find they can’t compete, they need to raise their game.
Calling yourself an “artist” is not a meal ticket for life. If people don’t want to pay you for the stuff you make, that’s not a fault in the people who aren’t buying it. You need to look closer to home for the answer to that one.
Just to clarify, I have zero issues with AI creating art, especially for a game. I don’t even mind it hanging in a gallery if it is clearly identified for what it is, but if it starts replacing art by humans, I draw a line there.
Yeah automated art factories will be just as soulless as most studio run focus group monstrosities, we must not turn that process into the art version of the infinite paper clip machine. Humanity might as well just end it there.
I would love to see how a true artificial intelligence would express it’s inner heart and thoughts. What mediuim would it choose? Or would it transcend all known human form of expression and create something we wouldnt even consider art?
“Don’t you see, I have displayed these one billion sprinkles in a random order that feels like a Monday to me.”
There is a belief amongst some that it’s impossible for humans to have a truly original idea. Everything we think is based on what we already know, and we extrapolate from there.
We can’t think of things totally outside our existing frame of reference. We don’t have the tools.
The current neural networks are trained to parrot us, so, not likely to come up with anything new soon. Maybe later versions.
I disagree with “no idea is new”. That’s like saying “this is not a new potato because it sprouted from another potato”.
I would have interpreted such a “not new!” claim as a provocative reminder to analyse my “new” ideas and become aware of the parallels to previous ideas that I have overlooked in my “inventor’s arrogance”. But putting old things together in new ways counts as new idea, in my book.
I tried to get chatgpt and dallee to recombine concepts in new ways and they couldn’t; or explain things to me using analogies, and the analogies were inconsistent or weak. So as far as I can tell, this added benefit isn’t happening yet.
There is no reason to assume that a software architecture modeled on our brains could do anything that would be conceptually impossible to our brains.
They could potentially do it many times faster, and involving more connections, and could thereby make leaps of thought in minutes that might take us years, and that would might seem like it’s more innovative, but they would still have to build one thing on top of the other. The process would not be inherently different.
Due to blazingly fast and more connected thought processes they would absolutely be more likely to arrive at brilliant solutions that might have taken humans decades to arrive at, though. This is mostly because the most brilliant ideas are mostly random affairs: The right mind must ponder the right things with the right information under the right circumstances to produce something truly brilliant. A very powerful AI would statistically have much better chances to hit those illusive criteria.
We’ve been seeing the effects of this already for some time now with genetic algorithms, which are not actually intelligent. They are built on the concept of random mutation and successive selection we know from evolution theory. These algorithms have come up with some eingineering solutions that have never occurred to any humans. But it’s not because they’re more creative. They’re not creative at all, no more than nature (which, given 14 billion years, has still managed to become the greatest creator of all, unless we want to get metaphysical about it).
So if we make a machine that is actually creative, and it can think through in a day what takes us a century, yes, the results are going to be pretty bloody astonishing, but conceptually it’s still subject to the same limitations.
Unless it uses that time to think up and implement a completely different brain architecture that we haven’t thought off. Now that would be a true singularity moment…
Plato said (in a very roundabout way) that those who seek power should not be given it, and that the only people suited to wield power are those who do not want it.
We can probably extend this principle: anybody who believes he can run the world better has demonstrated, purely by that belief, that they should never be allowed any role in running the world.
